CN103959332A - Image processing - Google Patents
Image processing Download PDFInfo
- Publication number
- CN103959332A CN103959332A CN201280056480.2A CN201280056480A CN103959332A CN 103959332 A CN103959332 A CN 103959332A CN 201280056480 A CN201280056480 A CN 201280056480A CN 103959332 A CN103959332 A CN 103959332A
- Authority
- CN
- China
- Prior art keywords
- image
- artifact
- parameter
- region
- detected region
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20004—Adaptive image processing
- G06T2207/20008—Globally adaptive
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Radiology & Medical Imaging (AREA)
- Medical Informatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
- Investigating Or Analysing Biological Materials (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
A system for processing an image comprises a region detector (1). The region detector comprises an artifact detector (7) for detecting a region comprising an artifact in the image. The system comprises a parameter determining unit (2) for determining a parameter, based on a portion of the image excluding the detected region. The system comprises an image processing module (3) for processing the image, using the derived parameter. The system comprises a display unit (5) for displaying the processed image with an indication of the detected region. The parameter determined by the parameter determining unit (2) can comprise a normalization parameter, and the image processing module (3) can be arranged for performing a normalization of the image according to the normalization parameter.
Description
Technical field
The present invention relates to image processing.
Background technology
In digital pathology, digital micro-analysis photo is made up of sample.Then by thereby produce digital picture present to virologist for assessment of.Watch traditional pathology of sample to compare with wherein directly passing through microscope, digital pathology have several advantages.First, digital pathology can increase handling capacity.This becomes more and more and is of great importance, because the working load of virologist in the face of increasing.Secondly, digital pathology can provide the means of the quality of improving diagnosis.Research shows, if use computerized algorithm to help virologist, the consistance of different pathological scholar's diagnosis can obviously be improved.
But digital pathology also can have several shortcomings.At present, one of shortcoming is that focal plane is fixed.Digital scanner typically only obtains an image of tissue.The adjustment of focal plane afterwards no longer may.This can produce out of focus artifact.But, also can obtain multiple different images with different focal planes.
Except this focus issues, other artifact conventionally appearing in pathology also appears in digital pathological image.Such artifact can comprise:
-pressure effect (pressure effect).
-fixing artifact (fixation artifacts) formation of used incorrect fixing agent, pollution, acid formalin haematine pigment (for example, due to).
-insufficient dehydration (for example, keeping being trapped in in-house water).
-dyeing artifact (staining artifacts) (for example,, due to uncleanly water-bath or inhomogeneous dyeing).
-by mistake appear at the object (for example, microorganism or hair) in image.
-defocus artifact (defocus artifacts).
-sensor noise.
Virologist rule of thumb can identify such artifact and ignore them in the time of interpretation of images.But automated image analysis program can have difficulty in pack processing aspect the image of one or more artifacts.
Current available numerical analysis technology is carried out for complete tissue or is at least detected for standardization and the core of complete visual field.
US2006/0014238A1 discloses the digital picture that strengthens immunohistochemistry compound and be coated to biological specimen on it; Remove the undesired unit of the predefined type in the digital picture strengthening and consider to avoid; Unit multiple interested in the digital picture that identification strengthens; One or more area-of-interests in the digital picture that identification strengthens; And remove the unit artifact in one or more identified area-of-interests to avoid considering, thereby automatically create the area-of-interest of one or more enhancings, for creating medical diagnosis or prognosis.
M.Macenko, M.Niethammer, J.S.Marron, D.Borland, J.T.Woosley, X.Guan, C.Schmitt and N.E.Thomas (hereinafter: the people such as Macenko) " A Method for Normalizing Histology Slides for Quantitative Analysis " is (in IEEE Int.Symposium on Biomedical Imaging, 2009) two kinds of mechanism of inconsistency of processing for overcoming dyeing are disclosed, thereby the wave carrier piece of processing under different condition or storing is taken in public standardization space to realize the quantitative test improving.This paper discloses and has automatically found the correct dyeing vector of image and then carry out the algorithm that color is deconvoluted.
In the image of other kind is processed, also can there is problem during containing the image of desired region not in pack processing.
Summary of the invention
The image processing with improvement will be favourable.In order to solve better this worry, a first aspect of the present invention provides a kind of system, comprising:
Area detector, comprises for detection of the artifact detecting device in region that comprises the artifact in image;
Parameter determining unit, determines parameter for the part of getting rid of beyond detected region based on described image; And
Image processing module, for processing described image by the parameter of deriving.
The advantage that described system has is: determine described parameter for the part of getting rid of in described image beyond detected region.Therefore, described parameter can be because detected region any do not expect or incoherent characteristic and distortion.Afterwards, process described image by the parameter of deriving.Therefore, with due to the attribute in detected region, the parameter of distortion is not carried out carries out image processing.In this manner, realize the improvement of the processing of image, because the dependent part based on described image assigns to determine described parameter.
Several regional areas of image, for example digital pathological image, may comprise artifact, and be dropped in order to analyze.Otherwise standardization and other image processing operations can be suboptimums.Therefore it is useful, in the time determining parameter, getting rid of the one or more regions that comprise artifact.Can process described image by this parameter subsequently.
Notice, in the people such as Macenko, in the time carrying out svd, for stability reasons, the pixel not dyeing is in fact determined threshold value.But such pixel of dyeing is not artifact, because some regions of pathology sample can by rights have little dyeing or not dyeing.
Digital Image Processing is benefited from concise and to the point strategy and is processed the scrambling in image.For example, in order for example to use computing machine to analyze (, core detection, classification, standardization etc.) to tissue, filtering artifact (for example, prepare by sample or introduce by scanning sequence) can be important to have reliable result.
Described system can comprise the display unit for showing processed image.And display unit can be arranged the instruction for showing the detected region that comprises artifact.This permission user sees in the time determining parameter, having got rid of which region.
Described system can comprise arranges for making user can control the user interface of the demonstration of the instruction of display unit to detected region.Can provide and allow to obtain the visual several options that are suitable for user.Under user's control, can generate a greater variety of visual so that allow user to described visual comparing.
Described user interface can be arranged for making user can open or close the instruction in detected region.This contributes to check rapidly the position in detected region, if desired, and in the time not needing the instruction in detected region, does not disturb view.In this manner, user can in the case of have and the instruction in not detected region watch processed image.User thereby can control described system to show individually processed image, and there is control system to show the possibility of processed image under the instruction with the region being excluded while determining described parameter.This allows the flexible use of system.
User interface can be arranged for making user can change the outward appearance of the instruction in detected region.This provides greater flexibility, and allows to adjust as required described instruction.For example, can make user between any one of following operation or combination, select: the color of selecting instruction, show the outer boundary in detected region, by giving prominence to and be detected region with particular color or with the detected region of specific pattern filling, or utilize and indicate this region such as the symbol of arrow.
Parameter determining unit can be arranged for get rid of settling the standard of the part parameter beyond detected region based on described image.Image processing module can be arranged the standardization for carry out carries out image according to normalizing parameter.Thereby can carry out standardization to whole image by the definite parameter of part based on getting rid of in described image beyond detected region.This degree of accuracy that allows to increase is determined parameter, still identical normalizing parameter is applied to whole image simultaneously.Can optimize this normalizing parameter for the part of getting rid of in described image beyond detected region, consistent standardization is still provided simultaneously.
Described system can comprise the mask unit of the mask for generating the detected region of instruction.Described parameter determining unit can be arranged for determining described parameter based on described mask.Specifically, described parameter determining unit can define and be excluded region with described mask.This be system effectively realize possibility.
Described image can comprise pathological image.Artifact appears in pathology regularly.Therefore, described system can advantageously be applied to pathological image.
Described image processing module can be arranged for also processing described detected region by the parameter of deriving.This allows the dependent part based on described image to assign to determine described parameter.Meanwhile, process described image by the parameter of identical derivation being also applied to detected region according to consistent mode.
Artifact detecting device can be arranged for detection of at least one the artifact comprising in following items: out of focus artifact, dyeing artifact, pressure artifact and fixing artifact.These are the important example that appear at the artifact in pathological image, can get rid of described artifact for the image processing parameter of these images by determining.
Described artifact detecting device can be arranged for detecting artifact by application detection technique.For example, described artifact can be had a specific appearance that can use method for checking object to detect by known.Such method for checking object itself is known for those of skill in the art.Such object detection technique can be for detection of and/or the effective means of description region.
On the other hand, the invention provides a kind of digital microscope that comprises set forth system.Such digital microscope may further include digital camera for obtaining image and for realizing processor and the storer of described area detector, parameter determining unit and image processing module.For example, digital microscope is digital pathology microscope.
On the other hand, the invention provides a kind of workstation that comprises set forth system.
Aspect another, the invention provides a kind of method of processing image, comprise
Detection comprises the region of the artifact in image;
Part based on getting rid of in described image beyond detected region is determined parameter; And
With derive parameter process described image.
On the other hand, the invention provides a kind of comprising for making processor system carry out the computer program of the instruction of the method for setting forth.
Those skilled in the art will recognize that, can combine two or more in above mentioned embodiment of the present invention, realization and/or aspect according to being considered to useful any mode.
Those skilled in the art will recognize that, technology disclosed herein can be applied to two dimension, three-dimensional and/or the view data of higher-dimension more.And, will recognize, technology disclosed herein can be applied to the view data obtaining from the Image Acquisition mode of photography, microscopy, computed tomography, x ray, magnetic resonance or any other kind.And, will recognize, technology described herein not only can be applied to pathological image and micro-image, but also is applied to the image of other type that comprises photographs and medical image.
On the basis of this description, those of ordinary skill in the art can carry out with the described amendment of described system and be out of shape amendment and the distortion of corresponding image acquiring device, workstation, system, method and/or computer program.
Brief description of the drawings
These and other aspect of the present invention makes an explanation becoming obviously and with reference to the embodiment describing hereinafter according to the embodiment hereinafter describing.
Fig. 1 is the block scheme for the treatment of the system of image.
Fig. 2 is the process flow diagram of processing the method for image.
Fig. 3 shows four examples of pathological image.
Fig. 4 shows the mask corresponding with the region that comprises artifact.
Fig. 5 shows IHC colored graph picture.
Fig. 6 shows the result of core Check processing.
Fig. 7 shows the result of core Check processing in the case of the instruction with detected region.
Embodiment
Can manually analyze pathological image.During manual analyzing, sample process artifact and organize artifact to be ignored by virologist, and by opening the focusing wheel on microscope, virologist can make the relevant portion of image keep focusing on.Recent trend instruction: pathology will be digitized in future soon.In the time using digital pathology, utilize computing machine to complete at least a portion of this analysis (for example core detects, core is classified, HER2-marks etc.).But known Computer Analysis may be experienced inaccurate result due to artifact and out of focus region.
By detecting first partly artifact and out of focus region and then abandoning these regions in analysis improves analysis result.This is not only applicable to pathological image, but also is applicable to the image of other type, comprises such as the medical image of computed tomography images, radioscopic image, magnetic resonance image (MRI) and the image of other kind.In medical image, the example of artifact is implant and reconstruct artifact.Other image such as digital photograph also may comprise artifact, for example compression artefacts, for example JPEG artifact.
Fig. 1 shows the aspect for the treatment of the system of image.Described system can utilize the computing machine of suitable programming or processor to realize at least in part.Alternatively, can provide special electronic circuit to realize any or all feature of described system.Described system can utilize workstation to realize.Described system also can comprise user interface hardware, and for example keyboard, mouse device, display and/or touch-sensitive display are controlled described system to receive input from user, for example, to initiate the processing operation about image.For example, and such user interface hardware can be for the result of display system, processed image.Described system can also comprise communication port in case with communicate such as the miscellaneous equipment in image acquiring device or digital picture warehouse.
Described system can comprise picture receiver 4.This picture receiver 4 is operably connected with communication port.Picture receiver 4 can be arranged for from such as comprising that another equipment for the server of the memory storage of image receives one or more images.Picture receiver 4 also can be arranged for directly communicating with image acquiring device (not shown).Described image can for example comprise the image of history image or another kind, describes as other in current description is local.
Described system can comprise area detector 1.Area detector 1 is arranged for detection of the region that meets specific criteria in image, more specifically, comprises the region of artifact.Such artifact can be organic object, for example insect or microorganism or hair.Artifact can be also artificial objects.The example of the artifact of mentioning in this paragraph can appear in pathological image.In pathology imaging, people are interested in the characteristic of tissue samples.It can be difficult observing these characteristics in the existing region of artifact in image.But artifact also can appear in other application.
The artifact of another kind is caused by the physical treatment of organizing.For example, if a part for tissue is under pressure, this part of this tissue can be out of shape, and for example, can destroy cellular construction.The artifact of another kind is caused by the microscopical optics setting such as focus and/or aperture.The artifact of another kind is caused by the digital pre-service of the image such as digital image compression.
Described system can comprise the artifact detecting device 7 for detection of the artifact in image.Artifact detecting device 7 can be exclusively used in the artifact that detects particular types.For example, artifact detecting device 7 can be arranged the insect that detects particular types for the shape based on insect and/or Identifying Technique of Object.Such technology is well known in the art itself.Described system can comprise multiple such artifact detecting devices 7, and wherein each artifact detecting device 7 can be arranged for detection of different types of artifact.For example, can detect different insects with different shapes.Can detect hair with another shape.And, can carry out individually modeling to the object of other kind.Different artifact detecting devices can be implemented as the artifact that detects other kind, and for example optical distortion and/or data and image processing are abnormal.Such artifact detecting device is well known in the art itself.Area detector 1 generates the expression in the region that comprises the one or more artifacts that detected by artifact detecting device 7.
Described system can comprise the parameter determining unit 2 of determining parameter for get rid of detected region part in addition based on image.Can assign to determine such parameter by the remainder of the mode analysis image that is known in the art according to itself.Determine that the mode of this parameter depends on the kind of image processing to be completed.Parameter determining unit 2 can operationally be coupled to area detector 1 to receive the expression comprising from the region of the artifact of area detector 1.Such expression can be the form of the coordinate of the bounding box of artifact.Other expression such as mask images is also possible.
Described system can comprise the image processing module 3 for process image by the parameter of deriving.For this reason, can utilize the suitable image processing operations of any expectation.For example, can carry out histogram equalization (using histogram vertical bar as parameter), can operative norm (using max pixel value as parameter), for example carry out object count by threshold value so that detected object, the wherein histogram of threshold value based on getting rid of the part beyond detected part in image.Based on the disclosure, in processing those skilled in the art's of other kind comprehension.
Described system can comprise the display unit 5 for showing processed image.Display unit 5 can (not processed) image original in showing and for showing processed image.And display unit 5 can be arranged the instruction for showing detected region.For example, can utilize region profile around to describe detected region, or can show detected region according to distinctive colors.Can be in not processed image and/or processed image the detected region of instruction.Display unit 5 can for example be embodied as the software unit of controlling external display device.
In addition, described system can comprise user interface 8.User interface 8 can be arranged and is used for making user can control described system.Except many other application, user interface 8 can also be arranged for making user can control the demonstration of the instruction of display unit to detected region.Described user interface can be configured to the artifact that makes user can manifest/show image and be dropped during the processing of described image.For example, can provide the button on graphic user interface to make user can open or close the instruction in detected region.In this manner, can in the case of the instruction in not detected region, show processed image and/not processed image.In the time that user selects to press button, can in the case of the instruction with detected region, show identical image.Such instruction can be indicated position and/or scope and/or the shape in detected region.And user interface 8 can be arranged for making user can change the outward appearance of the instruction in detected region.For example, can provide the list of the several options of user, user can be selected between described option.For example, user can select between the detected profile in region and the colouring in the coloured detected region of tool.Such colouring color can be completely opaque or translucent.Color and/or opacity are at user option.Can further make user can select the demonstration of the symbol of the position of indicating detected region.Such symbol can comprise arrow or point or any other icon.
For example, for above-mentioned purpose, user interface comprises one or more user interface elements.Such user interface element can comprise the member of graphic user interface.Also can use the user interface element such as any kind of hardware button.The instruction that user can thereby select he (or she) whether to want to be with or without detected region at tool, see processed image.Equally, can provide the user interface element that manifests original (not processed) image the instruction that makes user be with or without detected region at tool.The image processing of one type is standardization.Therefore, parameter determining unit 2 can be arranged for determining and will be used for standardized one or more parameter.Can be mean intensity, maximum intensity or histogrammic vertical bar value for the example of the parameter of normalizing operation.Image processing module 3 can be arranged the standardization for carry out carries out image according to one or more normalizing parameters.For example, can adjust strength level to change mean intensity or take the scope of used intensity to preset range by application correction factor.Histogram equalization is also possible.
Described system can comprise the mask unit 6 of the mask for generating the detected region of instruction.Such mask can be the form such as the image of binary picture, and wherein the first pixel value is detected the pixel of the part in region for conduct, and the second pixel value that is different from the first pixel value is not for being the pixel that is detected the part in region.Parameter determining unit 2 can arrange for determining parameter based on mask, and for example by only assessing, in not processed image, there is no masked image tagged be those pixels of the part in detected region.
Image processing module 3 can be arranged for also processing detected region by the parameter of deriving.So, although the value of parameter by parameter determining unit 2 part based on getting rid of beyond detected region in image determine, the parameter of the derivation that processing module can be based on whole image is carried out treatment step, to obtain the consistent processing of image.Alternatively, image processing module 3 can be arranged for only treatment step being applied to described image and get rid of the part beyond detected region.
Artifact detecting device 7 can be arranged for detection of out of focus artifact, dyeing artifact, pressure artifact and fixing artifact.Can use the different artifact detecting device 7 that is exclusively used in the artifact that detects particular types.
Artifact detecting device 7 can be arranged for detecting artifact by application detection technique.For example, shape or template matching technique can be for detection of the known artifacts having such as the particular community of given shape and/or color and/or texture.
Described system can be incorporated in digital microscope.Such microscope can have optical subsystem for amplifying area-of-interest and for the light signal obtaining from optical subsystem being converted to the digital camera of digital picture.Digital picture can be fed to picture receiver 4 for being processed according to the mode of setting forth by system herein.
Alternatively, described system can be merged in workstation.Such workstation can receive the image of being caught by digital microscope.
Described system is not limited to micro-image.Also can carry out optimization system for the image of other kind.
Fig. 2 shows the aspect of the method for processing image.The method comprises the step 201 of coming the region in detected image based on standard.This standard can be the existence of artifact in region.The scope in region can be accurately corresponding with the scope of detected artifact.Alternatively, the scope in region can be corresponding with the bounding box of the convex closure of artifact or artifact.Region can comprise artifact border around.These possibilities are also applicable to the region of being detected by area detector 1.After region being detected, the method proceeds to the step 202 of determining parameter based on the detected region of eliminating in described image part in addition.After determining parameter, the method proceeds to the step 203 of processing image by the parameter of deriving.Afterwards, the method can proceed to the step 204 that shows processed image on display device.Alternatively or in addition, described image can be stored on storage medium.The method can be included in computer program, and this computer program comprises for making processor system carry out the instruction of the method for setting forth.
In view of the description of the function of system as herein described, the method can be revised and/or be expanded by those of skill in the art.
Digitizing pathology have several advantages.First, it can increase handling capacity.This becomes more and more and is of great importance, because there is the increase of virologist's workload in the U.S., this is aging owing to population on the one hand, and on the other hand owing to a small amount of reduction in the virologist's of U.S.'s graduation quantity every year.Secondly, digital pathology provide the means of the quality of improving diagnosis.Research shows, if use computerized algorithm to help virologist, the consistance in virologist's diagnosis can be improved significantly.But digital pathology also have several shortcomings.Major defect is that focussing plane is fixed.Digital scanner typically only provides an image of tissue.Therefore the adjustment of focal plane is no longer possible.But, can record multiple images with different focal planes.
The not only artifact in the face of being introduced by digital scanner itself of numeral pathology.The example that possible appear at the artifact in the pathological image of any kind below:
-pressure effect.
-fixing artifact the formation of used incorrect fixing agent, pollutant and/or acid formalin haematine pigment (for example due to).
-insufficient dehydration (for example, keeping being trapped in in-house water).
-dyeing artifact (for example, due to uncleanly water-bath and inhomogeneous dyeing).
-ink artifact, for example, make the ink for mark surgical resection margins by healthcare provider.
At E.McInnes " Artifacts in histopathology " (in Comp Clin Path (2005) 13:100 – 108), the general introduction of the artifact in traditional pathology is disclosed.
In the time explaining pathology sample, virologist is trained ignores artifact by the numbers.But digital pathology system needs effective and concise and to the point strategy.For example, for the analysis (core detection, classification, standardization etc.) that uses computing machine to carry out tissue, any artifact of filtering (prepared and introduced by scanning sequence by sample) is important to improve reliability and/or the reproducibility of result.
As the example of possibility artifact, in the time of autopsy, pliers may be caused by the undue extruding of the pliers in flensing's hand due to flesh tissue the injury of lung before fixing.Acid formalin haematine pigment is to bear pigment at acid pH place with the dark-brown anisotropy microcrystalline iron producing that reacts of the protoheme of haemoglobin by the formic acid of the formalin from not containing buffering agent.There will be the shallow pigmented section in the white matter of spinal cord, because water may keep being trapped in tissue due to insufficient dehydration, the tissue that causes part not to be colored.Uncleanly water-bath also can comprise the residual fraction of prior structure, and this can become and merges in current organization section, for example, be present in the hepatic tissue in lung.The eosin thin slice of seeing in the focal plane of histotomy may be caused by the lake obtaining in the stock solution being never filtered.
Can complete standardization and core detection for complete visual field for setup action entirety or at least.But as instruction in the above example, several regional areas to be insecure or to be incoherent for analysis, and to be dropped in order analyzing.Otherwise standardization meeting is suboptimum, and can indicate false positive and/or false negative such as the result of the automatic analysis from kinetonucleus detection and/or automatic nuclear counting.In present practice, standard picture standardized technique comprises histogram equalization.
Where can determine partly artifact (similar out of focus and dyeing artifact).Such artifact can be dropped during the analysis of image.Artifact that can instruction is dropped during result visual of analyzing.
The artifact of several types can appear in the digitized image of tissue." Autofocusing Algorithm Selection in Computer Microscopy " (in the Intelligent Robots and Systems, 2005) of Y.Sun, S.Duthaler and B.J.Nelson discloses the technology that detects focusing and out of focus artifact.Can use derivative scheme, for example (amendment) Laplce or energy Laplce or based on statistics algorithm and, for example variance and autocorrelation technique.Correlation technique can detect dyeing artifact, because the blur spot that they can be used as dyeing liquid occurs.Also can detect fixing and pressure artifact.
After artifact being detected, can be by carrying out execution analysis by binary mask, indicate which pixel should be used for analyze and/or which pixel should not be used for analyze.The visual of the region that comprises artifact can complete by the relevant range in mask images for example.
Before image standardization, where can determine first partly artifact (for example out of focus and dyeing artifact).The artifact of finding can be dropped during the foundation of normalizing parameter.Standardization can only be applied to gets rid of artifact region in addition.Alternatively, standardization can be with based on getting rid of the region of artifact in image, definite normalizing parameter is applied to whole image.Alternatively, the artifact detecting can be for example by fade in technology or combine at the remainder coming with image that changes gradually that is excluded region internal standardization parameter of application, to avoid at the flip-flop that is excluded the boundary normalizing parameter between region and the remainder of image.
Fig. 3 has illustrated two immunohistochemistries (IHC) colored graph picture from identical sweet food dish (coupes).IHC is the known technology that can knit for analysis bank the unit of section.The picture left above looks like to show the first image that uses digital pathology scanner to obtain.Top right plot looks like to show the second image that uses digital pathology scanner to obtain.Lower-left image shows the first image after standardization.Bottom-right graph looks like to show standardization the second image afterwards.Fig. 4 illustrates the mask images for the first image.Black region in Fig. 4 with in the first image, wherein the image obtaining from digital pathology scanner, detect that the region of artifact is corresponding.Determine the normalizing parameter for the first image based on the first image obtaining with digital pathology scanner, but get rid of according to the masked region of the mask images shown in Fig. 4.With thus the normalizing parameter that produces image is carried out to standardization, and result is the lower-left image shown in Fig. 3.
Fig. 5 shows another IHC colored graph picture.Visible in image is dyeing pseudomorphism artifact 501 and multiple core 502.Fig. 6 shows the result of core Check processing.Blackening point is corresponding with the nuclear phase detecting.Before described processing, dyeing pseudomorphism artifact 501 is detected as detected region, and while this region during described processing, be left in the basket.Therefore, core do not detected at 601 places, position of dyeing pseudomorphism artifact 501.Fig. 7 shows the result identical with Fig. 6 of core Check processing.But Fig. 7 also shows the instruction 701 in detected region 501.In the example of Fig. 7, this instruction is the form of bar-shaped zone.The instruction of other kind, for example the detected profile in region or the use of different colours, be also possible.Described instruction thereby can indicate the scope in detected region.Alternatively, described instruction can comprise the symbol such as arrow, shows the position in this detected region the scope in the case of detected region is not shown.User interface 8 can make user between Fig. 6 (there is no instruction) and the demonstration of Fig. 7 (having instruction), switch.
To recognize, the present invention is also applicable to computer program, is particularly suitable for implementing being of the present inventionly arranged on carrier or the computer program of carrier.Described program can be the form of source code, object code, code intermediate source and the object code that partly compiles form or be suitable for any other the suitable form using in the realization of the method according to this invention.It should further be appreciated that, such program can have many different architecture Design.For example, the program code of realizing the function of the method according to this invention or system can be subdivided into one or more subroutines.Between these subroutines, many different modes of distribution function will be obvious for technician.Subroutine can be stored in together in an executable file to form complete program.Such executable file can comprise computer executable instructions, for example processor instruction and/or interpreter directive (for example Java interpreter directive).Alternatively, can or all be stored at least one external libraries file and static or dynamically link with master routine in working time the one or more of subroutine.Master routine comprises calling at least one times at least one subroutine.Subroutine can also comprise calling each other.The embodiment relevant to computer program comprises the computer executable instructions corresponding with each treatment step of at least one method of setting forth herein.These instructions can be subdivided into subroutine and/or be stored in can be by one or more files static or dynamically link.Another embodiment relevant to computer program comprise with herein set forth at least one system and/or each of product install corresponding computer executable instructions.These instructions can be subdivided into subroutine and/or be stored in can be by one or more files static or dynamically link.
The carrier of computer program can be any entity or the equipment of the program of can carrying.For example, described carrier can comprise storage medium, for example the magnetic recording media taking CD ROM or semiconductor ROM as the ROM of example or as an example of flash drive or hard disk example.And described carrier can be can transport vehicle, for example can be via cable or optical cable or the electricity or the optical signalling that transmit by radio or other means.When by program body now in such signal time, described carrier can be constructed by such cable or miscellaneous equipment or device.Alternatively, described carrier can be that program is embedded in to integrated circuit wherein, in the execution that this integrated circuit is suitable for carrying out correlation technique or being used in correlation technique.
Should be noted that above-described embodiment is illustrative rather than definitive thereof the present invention, and those of skill in the art can design many optional embodiment in the case of not departing from the scope of claims.In the claims, any Reference numeral being placed between round bracket should not be interpreted as limiting claim.The use that verb " comprises " do not get rid of except in the claims statement those elements or step element or the existence of step.Word " one " before element or " one " do not get rid of the existence of multiple such elements.The present invention can be by means of comprising the hardware of several different elements and realizing by means of the computing machine of suitably programming.In the equipment claim of enumerating several devices, several can the embodiment with identical items by of hardware in these devices.Unique fact is that some tolerance of enumerating in mutually different dependent claims is not indicated the combination that can not advantageously use these tolerance.
Claims (15)
1. for the treatment of a system for image, comprising:
Area detector (1), comprises the artifact detecting device (7) that comprises the region of artifact for detection of described image;
Parameter determining unit (2), determines parameter for the part of getting rid of beyond detected region based on described image; And
Image processing module (3), for processing described image by the parameter of deriving.
2. the system as claimed in claim 1, further comprises for showing the display unit (5) of processed image together with the instruction to described detected region.
3. system as claimed in claim 2, further comprises for making user can control the user interface (8) of the demonstration of the described instruction of described display unit to described detected region.
4. system as claimed in claim 3, wherein, described user interface (8) is arranged for making described user can open or close the described instruction in described detected region.
5. system as claimed in claim 3, wherein, described user interface (8) is arranged for making described user can change the outward appearance of the described instruction in described detected region.
6. the system as claimed in claim 1, wherein,
The described parameter of being determined by described parameter determining unit (2) comprises normalizing parameter; And
Described image processing module (3) is arranged the standardization for carry out described image according to described normalizing parameter.
7. the system as claimed in claim 1, further comprises the mask unit (6) for generating the mask of indicating described detected region, and wherein, described parameter determining unit (2) is arranged for determining described parameter based on described mask.
8. the system as claimed in claim 1, wherein, described image comprises pathological image.
9. the system as claimed in claim 1, wherein, described image processing module (3) is arranged for also processing described detected region by the parameter of described derivation.
10. the system as claimed in claim 1, wherein, described artifact detecting device (7) is arranged for detection of at least one in following items: out of focus artifact, dyeing artifact, pressure artifact and fixing artifact.
11. the system as claimed in claim 1, wherein, described artifact detecting device (7) is arranged for detecting described artifact by application detection technique.
12. 1 kinds comprise the digital microscope of the system as claimed in claim 1.
13. 1 kinds comprise the workstation of the system as claimed in claim 1.
14. 1 kinds of methods for the treatment of image, comprising:
Detect the region that described image comprises artifact;
Part based on getting rid of in described image beyond detected region is determined parameter; And
With derive parameter process described image.
15. 1 kinds of computer programs, comprise for making processor system carry out the instruction of method as claimed in claim 14.
Applications Claiming Priority (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US201161560831P | 2011-11-17 | 2011-11-17 | |
US61/560,831 | 2011-11-17 | ||
PCT/IB2012/056237 WO2013072816A1 (en) | 2011-11-17 | 2012-11-07 | Image processing |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103959332A true CN103959332A (en) | 2014-07-30 |
CN103959332B CN103959332B (en) | 2017-05-17 |
Family
ID=47358505
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201280056480.2A Active CN103959332B (en) | 2011-11-17 | 2012-11-07 | Image processing |
Country Status (6)
Country | Link |
---|---|
US (1) | US9916515B2 (en) |
EP (1) | EP2780888B1 (en) |
JP (1) | JP6342810B2 (en) |
CN (1) | CN103959332B (en) |
IN (1) | IN2014CN03657A (en) |
WO (1) | WO2013072816A1 (en) |
Families Citing this family (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2732432B1 (en) * | 2011-07-13 | 2020-01-08 | Koninklijke Philips N.V. | Method for automatically adjusting a focal plane of a digital pathology image |
WO2013072816A1 (en) | 2011-11-17 | 2013-05-23 | Koninklijke Philips Electronics N.V. | Image processing |
US10216762B2 (en) | 2014-06-04 | 2019-02-26 | Panasonic Corporation | Control method and non-transitory computer-readable recording medium for comparing medical images |
EP3251052B1 (en) * | 2015-01-31 | 2024-07-17 | Ventana Medical Systems, Inc. | Quality control of automated whole-slide analysis |
EP3424019A4 (en) * | 2016-03-01 | 2019-10-23 | Ventana Medical Systems, Inc. | Improved image analysis algorithms using control slides |
US10496893B2 (en) * | 2016-08-11 | 2019-12-03 | DiamondFox Enterprises, LLC | Handheld arthropod detection device |
US11715556B2 (en) | 2016-08-11 | 2023-08-01 | DiamondFox Enterprises, LLC | Handheld arthropod detection device |
US20190213736A1 (en) * | 2016-08-31 | 2019-07-11 | Koninklijke Philips N.V. | Apparatus for tubulus detection from a tissue biopsy |
DE102019217524A1 (en) * | 2019-11-13 | 2021-05-20 | Siemens Healthcare Gmbh | Method and image processing device for segmenting image data and computer program product |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101071474A (en) * | 2006-05-10 | 2007-11-14 | 航伟科技股份有限公司 | Body image abnormal area statistical detection method |
CN101689295A (en) * | 2007-06-29 | 2010-03-31 | 汤姆森许可贸易公司 | Apparatus and method for reducing artifacts in images |
US20100111441A1 (en) * | 2008-10-31 | 2010-05-06 | Nokia Corporation | Methods, components, arrangements, and computer program products for handling images |
CN102222318A (en) * | 2010-04-16 | 2011-10-19 | 深圳迈瑞生物医疗电子股份有限公司 | Grid artifact detection and suppression method and device |
Family Cites Families (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
AU670938B2 (en) | 1992-02-18 | 1996-08-08 | Neopath, Inc. | Method for identifying normal biomedical specimens |
US5978497A (en) | 1994-09-20 | 1999-11-02 | Neopath, Inc. | Apparatus for the identification of free-lying cells |
US20050136509A1 (en) | 2003-09-10 | 2005-06-23 | Bioimagene, Inc. | Method and system for quantitatively analyzing biological samples |
TWI245229B (en) | 2003-09-18 | 2005-12-11 | Ind Tech Res Inst | Region based illumination-normalization method and system |
JP4745080B2 (en) * | 2006-02-20 | 2011-08-10 | 猛 中浦 | X-ray diagnostic apparatus, image processing apparatus, and program |
US8019134B2 (en) | 2006-11-16 | 2011-09-13 | Definiens Ag | Automatic image analysis and quantification for fluorescence in situ hybridization |
EP2122571B1 (en) | 2006-12-19 | 2018-02-21 | Koninklijke Philips N.V. | Apparatus and method for indicating likely computer-detected false positives in medical imaging data |
JP2008165496A (en) | 2006-12-28 | 2008-07-17 | Toyota Central R&D Labs Inc | Image normalization device, object detection device, object detection system and program |
JP2009059326A (en) * | 2007-08-06 | 2009-03-19 | Nikon Corp | Imaging apparatus |
US9002134B2 (en) | 2009-04-17 | 2015-04-07 | Riverain Medical Group, Llc | Multi-scale image normalization and enhancement |
US8897601B2 (en) | 2009-09-28 | 2014-11-25 | Nec Corporation | Image conversion parameter calculation device, image conversion parameter calculation method, and program |
WO2013072816A1 (en) | 2011-11-17 | 2013-05-23 | Koninklijke Philips Electronics N.V. | Image processing |
-
2012
- 2012-11-07 WO PCT/IB2012/056237 patent/WO2013072816A1/en active Application Filing
- 2012-11-07 CN CN201280056480.2A patent/CN103959332B/en active Active
- 2012-11-07 JP JP2014541783A patent/JP6342810B2/en active Active
- 2012-11-07 US US14/359,187 patent/US9916515B2/en active Active
- 2012-11-07 IN IN3657CHN2014 patent/IN2014CN03657A/en unknown
- 2012-11-07 EP EP12801776.1A patent/EP2780888B1/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101071474A (en) * | 2006-05-10 | 2007-11-14 | 航伟科技股份有限公司 | Body image abnormal area statistical detection method |
CN101689295A (en) * | 2007-06-29 | 2010-03-31 | 汤姆森许可贸易公司 | Apparatus and method for reducing artifacts in images |
US20100111441A1 (en) * | 2008-10-31 | 2010-05-06 | Nokia Corporation | Methods, components, arrangements, and computer program products for handling images |
CN102222318A (en) * | 2010-04-16 | 2011-10-19 | 深圳迈瑞生物医疗电子股份有限公司 | Grid artifact detection and suppression method and device |
Non-Patent Citations (2)
Title |
---|
MARCO FORACCHIA ET AL.: "Luminosity and contrast normalization in retinal images", 《MEDICAL IMAGE ANALYSIS》, vol. 9, no. 3, 30 June 2005 (2005-06-30) * |
MARK W. POWELL ET AL.: "A Methodology for extracting objective color from images", 《IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS, PART B: CYBERNETICS》, vol. 34, no. 5, 31 October 2004 (2004-10-31) * |
Also Published As
Publication number | Publication date |
---|---|
US20140334711A1 (en) | 2014-11-13 |
WO2013072816A1 (en) | 2013-05-23 |
CN103959332B (en) | 2017-05-17 |
EP2780888A1 (en) | 2014-09-24 |
IN2014CN03657A (en) | 2015-10-16 |
US9916515B2 (en) | 2018-03-13 |
JP6342810B2 (en) | 2018-06-13 |
JP2014533857A (en) | 2014-12-15 |
EP2780888B1 (en) | 2019-08-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103959332A (en) | Image processing | |
JP7422825B2 (en) | Focus-weighted machine learning classifier error prediction for microscope slide images | |
CN107451998B (en) | Fundus image quality control method | |
Zaimi et al. | AxonSeg: open source software for axon and myelin segmentation and morphometric analysis | |
US8442280B2 (en) | Method and system for intelligent qualitative and quantitative analysis of digital radiography softcopy reading | |
US9076198B2 (en) | Information processing apparatus, information processing system, information processing method, program and recording medium | |
CN109791693A (en) | For providing the digital pathology system and related work process of visualization full slice image analysis | |
Hortinela et al. | Identification of abnormal red blood cells and diagnosing specific types of anemia using image processing and support vector machine | |
US11062168B2 (en) | Systems and methods of unmixing images with varying acquisition properties | |
CN105849274A (en) | Method and system for classifying and identifying individual cells in a microscopy image | |
Son et al. | Morphological change tracking of dendritic spines based on structural features | |
Joseph et al. | Quantitative and qualitative evaluation of deep learning automatic segmentations of corneal endothelial cell images of reduced image quality obtained following cornea transplant | |
JP2006507579A (en) | Histological evaluation of nuclear polymorphism | |
LEE et al. | Intensity correction of fluorescent confocal laser scanning microscope images by mean‐weight filtering | |
CN118284942A (en) | Digital analysis of pre-analytical factors in tissue for histological staining | |
CN110378949B (en) | Starch granule distribution analysis device and method thereof | |
US20190392612A1 (en) | Process For Determining The Infiltration Of Biological Cells In A Biological Object Of Interest | |
Smal | Particle filtering methods for subcellular motion analysis | |
EP3938999A1 (en) | Correlated image analysis for 3d biopsy | |
CN117597709A (en) | Method and apparatus for recording training data | |
Herold et al. | A machine learning based system for multichannel fluorescence analysis in pancreatic tissue bioimages | |
de Oliveira et al. | Segmentation of Cell Nuclei Regions in Epithelium of Prostate Glands. | |
Arias Vesga | Non-Nuclei Tissue Characterization of Histopathological Images: A Processing Step to Improve Nuclei Segmentation Methods |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |